The proposed training plan will integrate cognitive modeling and functional magnetic resonance imaging (fMRI) to investigate the neural correlates of decision biases and how they relate to biased processing of threat in high anxiety. Drift-diffusion models, which describe the processes underlying simple decisions, will be used to bridge behavioral and imaging data in service of identifying neural systems that correspond to different types of bias in decision making. These effects will be demonstrated across three different decision tasks to establish the generality of the decision components. The results are expected to demonstrate how drift diffusion models can dissociate decision biases that cannot be distinguished from the raw behavioral data. Further, the findings should show how fMRI data can be used to differentiate different types of bias that cannot be distinguished based on behavioral data or diffusion models alone. Thus cognitive modeling and fMRI will be used to augment traditional analyses of behavioral data and provide a deeper understanding of decision bias. The results of the first project will be used to provide useful information about th underlying sources of bias for threatening information that is typically observed in high anxiety. This endeavor will be useful to research in cognition in general, as it will provide guidance for researchers attempting to localize the source of behavioral effects from decision making tasks. The primary thrust of the training plan will be to employ this general approach to improve our understanding of the relationship between elevated levels of anxiety and biased processing of threatening information. Biased processing of threat is thought to underlie the etiology and maintenance of anxiety and mood disorders, thus disentangling the sources of cognitive bias has important implications for the identification and treatment of related psychopathologies. In particular, model-based fMRI analysis will be used to explore whether such biases reflect differences in the expectation of threat in the environment, differences in reward from identifying threat, or differences in top-down control of threat processing. The results will provide valuable insight into the nature of cognitive bias in high anxiety, and provide relevant information for clinicians looking to reduce this bias through cognitive training.